|Publication number||US4845538 A|
|Application number||US 07/152,702|
|Publication date||Jul 4, 1989|
|Filing date||Feb 5, 1988|
|Priority date||Feb 5, 1988|
|Publication number||07152702, 152702, US 4845538 A, US 4845538A, US-A-4845538, US4845538 A, US4845538A|
|Original Assignee||Emanuel Hazani|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (7), Referenced by (21), Classifications (28), Legal Events (7)|
|External Links: USPTO, USPTO Assignment, Espacenet|
1. Field of the Invention
The present invention relates generally to nonvolatile semiconductor read only memory (ROM) systems and, more particularly, relates to an improved electrically erasable programmable read only memory (E2 PROM).
2. Description of the Relevant Art
A great amount of research is being directed to improving E2 PROM memory systems. It is anticipated that these systems will eventually replace disk drives thereby substantially reducing the cost and read access time of storage in digital systems.
An E2 PROM cell is essentially an MOS transistor including source and drain diffusions coupled by a channel, a word line positioned over the channel, and a programmable floating gate positioned under the word line and above the channel. The cell stores a first binary state when the floating gate is charged with excess electrons and stores a second binary state when the floating gate is not charged with excess electrons.
During a read operation, the drain diffusion is charged to VD(R) and the word line is charged to VWL(R). If the floating gate is not charged, then the channel is inverted, current flows between the source and drain, and the voltage level of the drain diffusion decreases. This decrease is sensed to indicate that the second binary state is stored in the cell. If the floating gate is charged, then the threshold voltage of the MOS transistor is increased, the channel is not inverted, no current flows between the source and drain, and the voltage level of the drain diffusion does not decrease. This voltage level is sensed and indicates that the first binary state is stored in the cell.
Two methods utilized to charge the floating gate will now be described. The first method is to increase the potential difference between the source and drain diffusion to a level that induces avalanche breakdown at the drain junction. High energy electrons (hot electrons) will penetrate the oxide layer separating the floating gate from the channel and program the floating gate.
The second method is to cause electrons to tunnel from a grounded programming poly diffusion to the floating gate through a special tunnelling oxide layer. This tunnelling method requires three poly layers. The first layer is grounded and provides electrons during the charging operation, the second layer forms the floating gates, and the third layer forms the word lines. A select transistor must be included in each cell to allow discharging the the floating gate of a single cell in a row.
The discharging of the floating gates will now be described. In the hot electron embodiment, the word line is charged to a relatively high erase voltage level VWL(E) and the floating gate is capacitively coupled to a grounded drain electrode through the isolation oxide. Electrons then tunnel from the floating gate to the word line through a tunnelling oxide layer. Similarly, in the tunnelling embodiment, the first poly layer is grounded and the second poly layer is coupled to ground via high capacitive ratio between the layers. The word line is charged to VWL(E) and electrons tunnel from the floating gate to the word line.
Both types of systems have advantages and drawbacks. Turning first to the hot electron type of system, the structure only requires two poly layers and is thus easier to fabricate. However, hot electrons transferred via the oxide damage the oxide thereby reducing the number of programming cycles. Additionally, the production of hot electrons requires an off-chip power source capable of generating significant current. Further, the weak coupling of the floating gate to ground requires that the magnitude of the erase voltage be high. When arranged in an array matrix, selective cells cannot be eased without the erasure of the entire array.
Turning next to the tunnelling system, the program and erase operations are less stressful to the oxide and more cycles are possible. Additionally, less current is drawn than in the hot electron system. However, during the erase operation extraneous electrons tunnel from the first poly to the word line and thus cause excess stress to the oxide and higher than necessary current flow. Further, the three poly layer structure is complex and requires four control signal contacts, i.e., source, drain, word line, and select.
In view of the above, it is apparent that an improved E2 PROM system facilitating the use of an on-chip power supply, utilizes only two poly layers, and provides for a high number of program cycles is greatly required.
The present invention is an improved E2 PROM cell that utilizes electron tunnelling from the word line to the floating gate and includes only first and second poly layers. Only three control signals are required to read, program, or erase the cell.
In a preferred embodiment, the cells are formed along the major surface of a P- doped semiconductor substrate and arranged in a rectangular array at the intersections of vertical bit line diffusions and horizontal poly word lines. Each cell includes a modified first MOS transistor including a first channel region, disposed below the word line, connecting the source and drain diffusion. This first MOS transistor has a stacked poly gate with a floating gate having a first region underlying the word line and overlying a drain area of the channel of the transistor. The channel also includes a source area over which the floating gate is not disposed. The state of the bit stored in the cell depends on whether the floating gate is charged.
Each cell also includes a control region that is geometrically isolated from the source and drain diffusions. A second region of the floating gate overlies this control region so that the floating gate is strongly capacitively coupled to the control diffusion region. Additionally, the second region of the floating gate overlies a second channel region of a second transistor coupling the drain diffusion and control diffusion region. Further, this second region of the floating gate is adapted to be capacitively coupled to the drain duffusion.
According to one aspect of the invention, a selected cell is programmed by charging its drain diffusion to a programming voltage level and grounding its word line. The voltage level on the floating gate increases, due to the capacitive coupling to the drain and drain area of the channel, to invert the second channel region and couple the control diffusion region to the drain. Thus the control region is charged to programming voltage and this voltage is strongly coupled to the floating gate by the floating gate/control region capacitance.
According to a still further aspect of the invention, a selected cell is erased by grounding its drain and charging its word line to an erase voltage. A sufficient potential difference is generated to cause electrons to tunnel from the floating gate to the word line. For unselected cells in the same column, the word lines and drain are grounded. This voltage prevents the generation of a potential difference to cause tunnelling from the floating gate to the word line. Accordingly, a single cell in a column may be erased.
In one embodiment, a single cell is utilized as a circuit element that functions as an NMOS transistor having a programmable threshold voltage level. The three NMOS transistor control signals, VS, VD, and VG, are coupled to the source diffusion, drain diffusion, and P2 layer respectively.
For unselected cells in the same column, the word-lines are charged to an intermediate voltage to prevent the generation of sufficient potential difference to cause tunnelling from the word-lines to the floating gates. Accordingly, a single cell in a column may be erased.
The present cell causes very little stress to be placed on the oxide during the program and erase operations and can be reliably subjected to many such operations. As described above, tunnelling between the word line and floating gate is the least stressful method of programming and erasing. Further, extraneous current flow is eliminated due to the efficient geometry of the system.
Other features and advantages of the invention will become apparent in view of the drawings and following detailed description.
FIG. 1 is a top view of a preferred embodiment of the array;
FIG. 2 is a top view of a preferred embodiment of a cell;
FIGS. 3A-3D are cross-sectional views of the cell depicted in FIG. 2;
FIG. 4 is a schematic diagram of an electrical model of the cell;
FIG. 5 depicts an alternative structure for
FIG. 6 depicts a CMOS circuit including the cell of the present invention; and
FIG. 7 depicts a standard configuration of the circuit of FIG. 6.
FIG. 1 depicts the overall layout of the array of E2 PROM memory cells. Referring now to FIG. 1, the array is laid out on the major surface 10 of a P- doped monocrystalline silicon substrate. A first set of equispaced, vertical N+ regions 12 form the source/drain lines of the array. These source/drain lines are crossed by a first set of equispaced, horizontal polysilicon word lines 14. A second set of rectangular N+ control regions 16 are disposed between the source/drain regions 12 and geometrically isolated therefrom. A second set of floating gate poly layers 18 each include a first region disposed below an associated word line 14 and a second region disposed over an associated control region 16.
The word lines 14 and floating gates 18 are formed from separate poly layers that are deposited over the substrate and then selectively etched. Because the floating gates 18 are partially disposed below the word line 14, the layer utilized to form the floating gates is deposited first and the floating gates are referred to as poly 1 or P1. Similarly, the word lines are often referred to as poly 2 or P2. This terminology is utilized herein. A basic memory cell 20 is defined at the intersection of each source/drain line 12 and word line 14. The structure of the cell will now be described with reference to FIGS. 2 and 3A-3D.
In FIG. 2, the basic cell 20 is depicted with the drain line 12d positioned at the left and the source line 12s positioned at the right. Note that the line 12d functions as the source for the cell 20(1) disposed to the left of the cell depicted and the line 12s functions as the drain for the cell 20(r) disposed to the right of the cell depicted.
In FIG. 2, two transistors, T1 and T2, are depicted with both transistors having drain line 12d as their drains. Transistor T1 has the source line as its source, a first channel region 17 of the major surface underlying the word line 14 as its channel, and a stacked gate structure including the first region 18(1) of the floating gate 18 and the word line 14 as its control electrode. Transistor T2 has the control region 16 as its source, a second region 18(2) of the floating gate 18 as its gate, and a second channel region 19 of the major surface 10 disposed between the control region 16 and drain line 12d as its channel.
FIGS. 3A-3D are cross-sectional views taken along the cuts A-A', B-B', C-C', and D-D', respectively, of FIG. 2. FIG. 3A is a cross-sectional view of transistor T1. In FIG. 3A, the channel 17 of T1 includes a drain area DA disposed adjacent to the drain line 12d and below the first region 18(1) of the floating gate 18 and a source area SA disposed adjacent to the source line 12s and beneath the section of the word line 14 that does not overlay the floating gate 18.
During the program and erase operations electrons tunnel through the oxide layer 31 separating the first region 18(1) of the floating gate 18 and the word line 14. Generally, when a poly layer is deposited, the upper surface has small convex disks dispose thereon and the lower layer has small concave cups disposed thereon. It is well-known that the forward threshold required to induce tunnelling from a lower poly layer to an upper poly layer is lower than the reverse threshold voltage required to induce tunnelling from the upper poly layer to the lower layer.
FIG. 3B is a cross-sectional view of the transistor T2. A boundary of the floating gate 18 is adjacent to the drain line 12d and almost completely overlies the control region 16. The control region 16 and source line 12s are isolated by a field oxide region 32. FIGS. 3C and 3D depict the isolation of the N+ regions and channel regions by the field oxide layers.
FIG. 4 is an electrical schematic diagram illustrating the capacitive coupling of the floating gate P1 18 to the drain line 12d (CDP1), to the word line (P2) 14 (CP2P1), to the control diffusion 16 (CFNP1), and to the bulk semiconductor (CBKP1). The magnitude of these capacitances is determined by the area of overlap between P1 and the various circuit elements and the thickness of the oxide layer separating the structures.
In FIG. 4, the upper and lower plates of CDP1 are, respectively, the adjacent sections of the drain line 12d and the left edge of the floating gate 18. The left and right plates of CFNP1 are, respectively, the control region 16 and the section of the second region 18(2) overlying the control region. This coupling is much stronger than the drain/floating gate coupling because of the large overlap and the thinner oxide thickness. The upper and lower plates of CP2P1 are, respectively, the overlapping regions of the word line 14 and first region 18(1). Finally, the lower and upper plates of CBKP1 are, respectively, the bulk area of the channel of T1 and the first region 18(1) of the floating gate.
The following equation expresses the floating gate voltage (VP1) in terms of the above-described capacitances, the charge (QP1) on the floating gate, and the voltage levels on the wordline (VP2), the control electrode (VFN), the drain (VD), and the bulk semiconductor substrate (VBK): ##EQU1##
The program and erase operations will now be described for an embodiment having the dimensions shown on FIG. 2 and having the following oxide thicknesses:
TP2P1 =600 angstroms
TFNP1 =250 angstroms
TDP1 =400 angstroms
TDAP1 =400 angstroms
For this embodiment the forward tunnelling threshold voltage level is about 8 volts, the reverse tunnelling threshold voltage level is about 12 volts, and the threshold voltage of T1 and T2 is about 0.8 volts.
Table 1 lists the voltages on the drain 12d, source 12s, and word lines 14 for programming a selected cell to store the first binary state (floating gate charged) and the same voltage states for all unselected cells. Table 1 also list the value of the current (IDS) that flows from source to drain in T1. It is critical that unselected cells not be programmed by tunneling, or have a parasitic IDS current that might cause hot electron programming.
TABLE 1______________________________________ V D V S V P2 IDS______________________________________Selected 15 8 0 0Unselected 8 8 0 0On Same RowUnselected On 15 8 8 0Same Columns______________________________________
The programming of the selected cell will now be described with reference to Table 1 and FIG. 4. The drain voltage, VD, is coupled to the floating gate 8 by CDP1 and the voltage on the floating gate rises to about 2 volts which is higher then the VT of T1. Thus the drain area DA of the channel is inverted and acts as an additional capacitor plate to couple the drain voltage to P1. This capacitor is actually CBKP1 in FIG. 4. The source area SA of the channel is not depleted because T1 is off due to the fact that WL 14 is held at OV. The coupling via CBKP1 brings the potential of P1 to above 2 volts and is sufficient to invert and deplete the channel of T2. Thus T2 is turned on to charge the lower plate (the control region 16) of CFNP1 to VD minus the VT of T2 (0.8V).
This voltage on the control electrode is strongly coupled to the floating gate by CFNP1. Accordingly, the voltage levels VFN and VB in equation 1 are equal to VD -VT2 and 15V respectively. Substitution of these values into equation 1 with QP1 =0 (floating gate not charged) and VP2 = (word line at ground) indicates that the magnitude of VP1 is greater than 12 volts. Because P2 is grounded, this voltage on P1 is applied across the electrodes of CP1P2 and exceeds the reverse threshold voltage of the oxide layer so electrons tunnel from P2 to P1.
For unselected cells in the same row, VD is 8 volts and VP1 is about 6 volts. This voltage difference is insufficient to cause reverse tunnelling. For unselected cells in the same column, the potential of the word lines 14 is at 8V, thus the potential difference between the floating gates 18 to the control gates 14 is about 5 volts, which is below the reverse or forward tunneling threshold. Thus these cells will not program undesirably due to tunnelling of electrons. In addition VGS of T1 is 0 volts, because VP2 is 0 volts and SA is not depleted, so that no IDS exists. Thus accidental hot electron programming will not occur. Accordingly, the present system provides for programming a single cell utilizing only three control signals, VD, VS, and VP2.
Thus, the coupling between the drain 12d and the floating gate 18 is sufficient to turn on T2 which, in turn, facilitates charging the control region 16 to almost the full programming voltage. From eq. 1, this high value of VFN multiplied by the large value of CFNP1 results in a high value of VP1 and facilitates the use of lower programming voltage. Additionally, very little current is drawn because the floating gate is charged by the tunnelling mechanism.
Table 2 lists the voltage levels on the drain, source, and word lines for erasing (discharging) a selected cell to erase the first binary state (floating gate charged) and the values of these voltage levels for all unselected cells. It is critical that unselected cells on the same column and different rows not be erased.
TABLE 2______________________________________ V D V S V P2______________________________________Selected 0 0 15Unselected in same column 0 0 0______________________________________
For the selected cell, VD is at ground and VP2 is 15 volts. In this case, equation 1 indicates that the potential difference across CP2P1 is greater than the forward tunnelling voltage so electrons tunnel from the floating gate to the word line to erase (discharge) the floating gate. For unselected cells in the same column VP2 is 0 volts. In this case equation 1 indicates that the potential difference across CP2P1 is not sufficient to cause forward tunnelling across the oxide layer. Accordingly, a single cell in a column may be erased.
An alternative structure for CFNP1 is depicted in FIG. 5. The layout of cell 20 is as depicted in FIG. 2, however, the structure along cut B-B' is as depicted in FIG. 5 instead of FIG. 2A.
Referring now to FIG. 5, the control region 16 and second region 18(2) of the floating gate are implemented as a trench structure.
The process for forming this structure is well known and is utilized in the fabrication of Megabit DRAMs. The advantages of utilizing this structure are twofold. First, the footprint of the CFNP1 will decrease thereby reducing the area of the cell 20. Secondly, the magnitude of CFNP1 will increase because the trench structure provides for more overlap between the layers.
Because the present cell requires only three control signals, VD, VS, and VP1, it operation is analogous to a standard NMOS transistor with VG =VP1. However, the present cell has the additional advantage of having a programmable threshold voltage. Accordingly, the cell may be advantageously incorporated into NMOS or CMOS logic arrays, gate-arrays, and other forms of programmable logic.
If the cell is utilized as a discrete circuit element then the long source and drain diffusions 12s and 12d of FIG. 2 are replaced by isolated source and drain diffusion regions and the word line 14 is replaced by an isolated control gate.
FIG. 6 is a circuit diagram of a CMOS invertor having a high-impedance output capability. Referring to FIG. 6, a PMOS transistor 60 has its source terminal coupled to a VC VP input terminal 62 and its drain terminal coupled to the drain diffusion 12d of present cell 20. The source diffusion 12s of the cell 20 is coupled to a VSS terminal 64. The P2 layer 14 of cell 20 and the gate of the PMOS transistor 60 are coupled to an input terminal 66. An output terminal 68 is connected to the coupled drain terminals.
In normal operation the floating gate of the cell 20 is not charged and the threshold voltage of the cell 20 is less than 5 volts. Accordingly, when Vin =5 volts the cell 20 conducts and the PMOS transistor 60 is off so that VSS is coupled to the output terminal 68 thereby inverting the input signal. When Vin =0 volts, the PMOS transistor 60 is on, the cell 20 does not conduct and VC VP is coupled to the output terminal 68 thereby inverting the input.
To convert to the high impedance mode the floating gate 18 of the cell is charged to increase the threshold voltage of the cell to greater than 5 volts and Vin is set to 5 volts. Thus, both the cell 20 and the PMOS transistor 60 are nonconducting thereby isolating the output terminal 68 from the input terminal 60, the VC VP terminal 62, and the VSS terminal 64.
This circuit is particularly useful when the output terminals of several invertor circuits are coupled to a signal bus and it is desired to transfer information from only one of the invertors to the bus. The cells in the non-selected invertor circuits are programmed and the input voltage in held at 5 volts.
The programming or erase operations are as described above. To program the cell 20 in the circuit of FIG. 6, Vin is set to 0 volts and VC VP set to 15 volts. The drain 12d of the cell 20 goes to 15 volts because the PMOS transistor 60 is on. Additionally, since the source 12s and P2 layer are both at 0 volts no current flows through the cell 20. Under these bias conditions, the floating gate 18 is charged and the cell 20 is programmed to a high threshold voltage. To erase a programmed cell 20, Vin is pulled to 15 volts and VC VP is set at 5 volts. In this case, the PMOS transistor 60 is off and the output terminal 68 is coupled to the VSS terminal 64 through the cell 20. Under these conditions the floating gate 18 will discharge
FIG. 7 is a circuit diagram of a standard CMOS circuit for implementing a CMOS invertor with a high-impedance output feature. An extra select transistor 80 is required. If the SELECT signal is low then the output is isolated and if the SELECT signal is high the circuit functions as an invertor. Several advantages of the circuit utilizing the present cell 20 (FIG. 6) are now apparent. It requires one less transistor and one less control signal and is thus easier to manufacture. Further, because of the nonvolatile nature of the cell, the circuit of FIG. 6 remains programmed in the event that power is interrupted. The circuit of FIG. 7 must be reprogrammed using special software routines.
The invention has now been described with reference to several embodiments. Substitutions and alternative will now be apparent to persons of ordinary skill in the art. In particular, the doping level along the boundary of the floating gate adjacent the drain diffusion can be enhanced to increase the capacitive coupling between the drain and the floating gate. Alternatively, this boundary could slightly overlap the drain diffusion to increase this coupling. Further, if desired, it is possible to charge the floating gate utilizing the hot electron method by supplying a signal to the word line during the programming operation. Accordingly, it is not intended that the invention be limited except as provided by the appended claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4402064 *||Nov 25, 1981||Aug 30, 1983||Fujitsu Limited||Nonvolatile memory|
|US4442447 *||Mar 7, 1983||Apr 10, 1984||Rca Corporation||Electrically alterable nonvolatile floating gate memory device|
|US4477883 *||Feb 18, 1982||Oct 16, 1984||Tokyo Shibaura Denki Kabushiki Kaisha||Electrically erasable programmable read only memory|
|US4513397 *||Dec 10, 1982||Apr 23, 1985||Rca Corporation||Electrically alterable, nonvolatile floating gate memory device|
|US4580247 *||May 1, 1985||Apr 1, 1986||Deutsche Itt Industries Gmbh||Semiconductor floating gate memory cell|
|US4628487 *||Aug 14, 1984||Dec 9, 1986||Texas Instruments Incorporated||Dual slope, feedback controlled, EEPROM programming|
|US4663740 *||Jul 1, 1985||May 5, 1987||Silicon Macrosystems Incorporated||High speed eprom cell and array|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US5006918 *||Jul 24, 1989||Apr 9, 1991||International Business Machines Corporation||Floating orthogonal line structure for X-Y wiring planes|
|US5019881 *||Aug 10, 1989||May 28, 1991||Samsung Electronics Co., Ltd.||Nonvolatile semiconductor memory component|
|US5087583 *||Jul 27, 1989||Feb 11, 1992||Emanuel Hazani||Process for EEPROM cell structure and architecture with shared programming and erase terminals|
|US5099297 *||Jul 10, 1989||Mar 24, 1992||Emanuel Hazani||EEPROM cell structure and architecture with programming and erase terminals shared between several cells|
|US5166904 *||Feb 10, 1992||Nov 24, 1992||Emanuel Hazani||EEPROM cell structure and architecture with increased capacitance and with programming and erase terminals shared between several cells|
|US5247346 *||Mar 29, 1991||Sep 21, 1993||Emanuel Hazani||E2 PROM cell array including single charge emitting means per row|
|US5278785 *||Jun 12, 1991||Jan 11, 1994||Emanuel Hazani||Non-volatile memory circuits and architecture|
|US5303185 *||Feb 10, 1992||Apr 12, 1994||Emanuel Hazani||EEPROM cell structure and architecture with increased capacitance and with programming and erase terminals shared between several cells|
|US5332914 *||Oct 17, 1991||Jul 26, 1994||Emanuel Hazani||EEPROM cell structure and architecture with increased capacitance and with programming and erase terminals shared between several cells|
|US5504706 *||Oct 12, 1993||Apr 2, 1996||Texas Instruments Incorporated||Low voltage Fowler-Nordheim flash EEPROM memory array utilizing single level poly cells|
|US5784327 *||Jul 30, 1997||Jul 21, 1998||Hazani; Emanuel||Memory cell array selection circuits|
|US5930619 *||Feb 8, 1996||Jul 27, 1999||International Business Machines Corporation||Method of making trench EPROM simultaneously with forming a DRAM cell|
|US5932908 *||Dec 11, 1996||Aug 3, 1999||International Business Machines Corporation||Trench EPROM|
|US6136652 *||Jul 9, 1997||Oct 24, 2000||Hazani; Emanuel||Preventing dielectric thickening over a channel area of a split-gate transistor|
|US6236082 *||Aug 13, 1998||May 22, 2001||National Semiconductor Corporation||Floating gate semiconductor device with reduced erase voltage|
|US6368917||Nov 21, 2000||Apr 9, 2002||National Semiconductor Corporation||Methods of fabricating floating gate semiconductor device with reduced erase voltage|
|US6403455||Aug 31, 2000||Jun 11, 2002||Samsung Austin Semiconductor, L.P.||Methods of fabricating a memory device|
|US6689668||Aug 31, 2000||Feb 10, 2004||Samsung Austin Semiconductor, L.P.||Methods to improve density and uniformity of hemispherical grain silicon layers|
|US7075127||Jan 29, 2004||Jul 11, 2006||Infineon Technologies Ag||Single-poly 2-transistor based fuse element|
|US20050167728 *||Jan 29, 2004||Aug 4, 2005||Chandrasekharan Kothandaraman||Single-poly 2-transistor based fuse element|
|WO2005076357A1 *||Nov 26, 2004||Aug 18, 2005||Infineon Technologies Ag||Single-poly 2-transistor based fuse element|
|U.S. Classification||257/318, 365/185.31, 257/317, 257/E27.103, 257/E29.304, 365/185.02, 365/185.28|
|International Classification||H01L29/788, G11C16/10, G11C16/04, G11C11/56, H01L27/115|
|Cooperative Classification||G11C11/5621, G11C16/0425, G11C16/10, G11C11/5628, H01L27/115, G11C2029/0403, G11C2211/5611, G11C11/5635, H01L29/7883|
|European Classification||G11C11/56D2, G11C16/10, G11C11/56D2E, G11C11/56D, G11C16/04F2, H01L27/115, H01L29/788B4|
|Dec 24, 1992||FPAY||Fee payment|
Year of fee payment: 4
|Feb 11, 1997||REMI||Maintenance fee reminder mailed|
|Jun 12, 1997||FPAY||Fee payment|
Year of fee payment: 8
|Jun 12, 1997||SULP||Surcharge for late payment|
|Jan 23, 2001||REMI||Maintenance fee reminder mailed|
|Jul 1, 2001||LAPS||Lapse for failure to pay maintenance fees|
|Sep 4, 2001||FP||Expired due to failure to pay maintenance fee|
Effective date: 20010704